Immunomodulator and Biologic Combination Therapy in IBD: The Debate That Just Won’t Go Away?

More than two decades after the first case reports of infliximab therapy in Crohn’s disease [CD], readers of the Journal of Crohn’s and Colitis might expect a complete understanding of the optimal use of anti-tumour necrosis factor [TNF] therapies in inflammatory bowel disease [IBD]. Yet in 2020, we still see reports of studies addressing topics including optimal dose selection and dose monitoring. One apparently perennial debate is whether to combine anti-TNF drugs with adjunctive immunosuppression. Post-hoc analysis of initial registration trials did not show any difference in outcomes stratified by baseline immunosuppression.1 Growing awareness of the risks of thiopurines focused attention on thiopurine withdrawal. In 2008, Gert van Assche and colleagues showed that patients with CD in remission on infliximab and azathioprine were no more likely to flare over 2 years of follow-up if randomised to azathioprine cessation.2 In guidelines issued the following year, the American College of Gastroenterology [ACG] gave a cautious recommendation favouring maintenance treatment with anti-TNF monotherapy. Then, the 2010 SONIC study showed that patients with CD who were naïve to both anti-TNF and thiopurine therapy did better on the combination of both than on either drug alone.3 This was for infliximab and azathioprine and CD—what of adalimumab, of methotrexate, and of UC? Subsequent randomised trials have addressed some of these permutations. Notably, for UC, infliximab, and azathioprine, the UC SUCCESS trial showed combination therapy was superior.4 Conversely for CD, adalimumab, and azathioprine, and for CD, infliximab and methotrexate, respectively, the DIAMOND and COMMIT studies showed no apparent benefit to combination therapy.5,6 Based on pharmacokinetic data, a narrative emerged: infliximab is more immunogenic that adalimumab, and therefore the addition of immunosuppression confers benefit, reflected in higher drug levels. For this narrative, COMMIT represents an inconvenience explained away by pointing to high levels of concomitant steroid usage and the observations that trough infliximab levels were nonetheless numerically higher in the combination therapy arm. This narrative has been supported in the prospective observational UK PANTS study, confirming that anti-drug antibody formation was indeed more common with infliximab than adalimumab and was reduced by concomitant immunosuppression.7 But some observations in PANTS are discordant with trial data: whereas the absolute risk of anti-drug antibody formation was lower for adalimumab than infliximab, the relative risk reduction conferred for both drugs through the addition of an immunosuppressant was similar, suggesting that a larger or longer study than DIAMOND might have shown benefit to combination therapy for adalimumab. Likewise, the effects on anti-drug antibody formation in PANTS were similar for thiopurines and methotrexate, suggesting that COMMIT might indeed have been an anomaly. In this edition of the Journal, Targownik and colleagues report their analysis of outcomes for 11 244 Canadian patients with IBD receiving treatment with anti-TNF therapy.8 Using data from four population-level health care databases, the authors tracked a composite, surrogate marker of treatment failure [unplanned IBD-related hospitalisation, IBD-related resectional surgery, corticosteroid prescription, biologic switch] and used a multivariable analysis to model the effects of choice of biologic and of immunosuppressant. In both CD and UC, and with infliximab or adalimumab, patients receiving combination therapy had better outcomes. The choice of immunomodulator did not show a significant effect in CD, but better outcomes were seen with azathioprine than methotrexate in UC. The authors note that their findings would support more widespread use of combination therapy, but that overall rates of combination therapy nevertheless remained low across all permutations of disease and drugs. How can we synthesise these findings in clinical practice? We note first that scrutiny of risks of combination therapy remains appropriate, but in most IBD populations the absolute risk of complications with combination therapy remain low and must be set against the risks of treatment failure. Second, whereas DIAMOND provided randomised controlled trial [RCT]-level evidence against the need for combination therapy with adalimumab, PANTS demonstrated a clear benefit to combination therapy in terms of reduced immunogenicity, and Targownik and colleagues now demonstrate improved clinical outcomes. Although we are mindful of the hierarchies of evidence here, the ‘real-world’ studies with all their limitations nonetheless provide support for the idea that there are indeed medium- to long-term benefits still to be had with using adalimumab in combination therapy. Third, we do not favour the preferential use of methotrexate over a thiopurine as a combination therapy. Although methotrexate is associated with lower risks of opportunistic infections and malignancies, the findings of the present study suggest that, in UC at least, methotrexate is the less effective agent. Although benefits conferred by combination therapy may relate more to pharmacokinetic effects, these findings also fit with negative trial data for methotrexate monotherapy in UC. What about methotrexate in CD? Here the situation is more subtle: the present study showed overall benefit for patients on combination therapy, but a non-significant trend towards worse outcomes with methotrexate than with azathioprine. COMMIT failed to demonstrate a benefit of methotrexate combination, but then again, PANTS suggested that methotrexate was equally effective in reducing the risks of anti-drug antibody formation. Nevertheless, clinical treatment failure with anti-TNF therapy can and does occur in the absence of anti-drug antibodies. Where does this leave us? The pragmatic approach is to consider previous and current treatment history as well as patient-stratified risks of thiopurine-related complications at the point of starting an anti-TNF; then, if opting for combination therapy, to continue any current immunomodulator providing it is well tolerated or, where appropriate, to offer a thiopurine to the immunomodulator-naïve. As we develop the next generation of IBD therapeutics, we must avoid finding ourselves having the same conversations two decades from now, by building into trial programmes appropriate mechanistic, pharmacokinetic, and subgroup analyses. For example, the observation that post-hoc analysis has shown similar clinical outcomes for CD patients treated with or without concomitant immunomodulators, during the UNITI trials of ustekinumab, may be reminiscent of the observations made in the early days of anti-TNF therapy. However, the finding that there was no effect on pharmacokinetics of ustekinumab with concomitant immunosuppression is powerful and provides more reassurance.9 Finally, we acknowledge that the rift between the purity of clinical trial data and the pragmatism of real-world observational studies, such as the work of Targownik, poses clear problems for clinical guideline writers in an era where we are emphasising quality of evidence and rigour in guideline preparation. The recent ECCO and ACG CD guidelines both give strong guidance on what to do with patients who fit the profile of those included in the clinical trials, which all recruited patients naïve to both anti-TNF therapy and to immunosuppressants and measured outcomes at relatively early time points. Nevertheless, most patients will fall outside this category and require long-term treatment. Here guidelines must, necessarily, reflect the uncertainty and lower levels of evidence available. It seems unlikely that one definitive RCT [or set of RCTs] can or should now be designed to close down this debate, but rather that sophisticated and large-scale registry studies, such as the present study, will interrogate our past and so provide guidance for our future. No relevant funding was received. TR declares research/educational grants and/or speaker/consultation fees from Abbvie, BMS, Celgene, Ferring, Gilead, GSK, LabGenius, Janssen, Mylan, MSD, Novartis, Pfizer, Sandoz, Takeda, and UCB, all outside the submitted work. NAK declares personal fees from Falk, Takeda, and Pharmacosmos; other fees from Janssen; and non-financial support from Janssen, AbbVie, and Celltrion, outside the submitted work. TR and NAK jointly drafted and prepared the manuscript.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.